技能雷达 · GITHUB
值得安装的 Agent Skills。
汇总公开仓库中已验证的 Skills。安装前先看清它能做什么、包内包含什么、热度如何，以及来源是否可靠。
- 目录范围
- 已验证
- Skills
- 5,771
- 仓库
- 824
- 最近同步
- 2026年9月8日
robsonrunginterview-me
Settle feature and design choices through interview rounds of up to five questions. Use when the user asks to be interviewed or to settle requirements before a PRD; produce a decision record for to-prd.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungknowledge-graph
Build and query a document knowledge graph with traceable evidence. Use for multi-hop answers, corpus-wide themes, or shared graph memory; use direct retrieval for single-document or single-hop questions.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungmacro-architecture
Choose a system architecture style or assess service decomposition and data ownership. Use for monolith, service, event-driven, or distributed-system boundaries; use architecture-lens for code-level structure.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungmodels-consensus
Run a read-only council of model opinions after the user approves its model plan. Use only when the user invokes this skill or explicitly asks for extra model opinions; never use as an automatic escalation.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungopen-pr
Commit scoped changes, push a branch, and open or update a GitHub pull request. Use when the user asks to open a pull request, commit and push work, or write or update a pull request description. Do not use to implement work or merge a pull request.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungopencode-runner
Guide OpenCode CLI runs through the host agent approval flow. Use when users explicitly request OpenCode, want to compare OpenCode output with another model, or need an OpenCode perspective without a bundled runner script.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungpeer-sessions
Coordinate peer sessions through native delegation or a durable mailbox. Use for a requested peer fleet, cross-session handoffs, or resumable replies; use models-consensus for a requested opinion council.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungpi-runner
Run a prompt through Pi CLI with provider and model pinned per call. Use when Pi is requested or an approved workflow selects a Pi route, including the Kimi, GLM, Qwen, and Gemma seats.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungresolve-pr-feedback
Evaluate pull request feedback against the code, apply valid fixes, and reply to or resolve the addressed threads. Use only when the user asks to address pull request feedback or review threads. Do not use to find new issues in a diff.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungsafe-incremental-coding
Build characterization tests before a risky change to untested legacy code. Use to preserve existing behavior while making the affected code testable, then hand behavior changes to tdd.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungsecurity-gate
Identify security decisions during an interview and classify slices for deep or standard review. Use when defining an exposed security surface or planning its review; full-review performs the code review.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungsession-handoff
Create an immutable session handoff or orient from a selected handoff. Use when the user asks to hand off current work, save continuity notes, resume from a handoff, or inspect available handoffs. Do not use for an ordinary session summary.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungskill-expert
Create, audit, improve, or package SKILL.md workflows and their supporting resources. Use for skill authoring, discovery problems, or workflow simplification. For project instruction files, use agents-md-craft.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungsoftware-design-philosophy
Reduce design complexity through deep modules, information hiding, and coherent interfaces. Use when a module or API is hard to understand or change; use clean-code for local cleanup.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungsummarize
Create a concise continuity summary from the active session. Use when the user asks to summarize current work, prepare a handoff, or compact session context. Do not use to summarize an unrelated document or to store a handoff.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungtdd
Implement behavior through the red-green-refactor loop. Use when test-first execution is requested or selected by the coding workflow; use safe-incremental-coding first for untested legacy code.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungtest-lens
Judge test value and choose what to test or mock. Use for brittle tests, test reviews, and unit or integration test design; use tdd for the test-first execution loop.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungto-prd
Turn a settled decision record into a PRD for user approval. Use after interview-me or a settled product discussion; to-tasks consumes the approved PRD.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungto-prototype
Build a throwaway experiment to answer a design question that reading cannot settle. Use for technical spikes, state-model demos, or visual alternatives; use a tracer bullet when the code must ship.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungto-tasks
Break an approved PRD into reviewable vertical slices and publish the approved task queue. Use when planning implementation tasks; implement-tasks handles model approval and execution.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungui-ux-pro-max
Choose a reusable UI design system or review cross-screen UX using a design database. Use for visual systems, interaction rules, or UI quality; use frontend-design for bespoke visual implementation.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungverify-changes
Run a repository's deterministic checks from its own command surface. Use when the user asks to verify a branch, run repository checks, or prove that build and tests pass. It captures command evidence, does not repair code, and does not replace browser testing.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
robsonrungworktree
Set up an isolated git worktree for new work or an existing ref. Use when the user requests isolation or an implementation plan requires separate workspaces; reuse suitable existing isolation.
- 安装量
- 2
- GitHub Stars
- 2
- 最近更新
- 9月8日
apidojo-ioanalyzing-competitor-tiktok-content-strategy
Analyzes a competitor's TikTok content strategy and top-performing videos using apidojo's scrapers on Apify. Triggers when the user asks to: see what content a competitor posts on TikTok, analyze which TikTok videos perform best for a brand, reverse-engineer a competitor's TikTok strategy, find out how often a competitor posts and what gets the most engagement, study a brand's TikTok presence, compare TikTok content strategies across competing brands, or find out what topics a competitor covers on TikTok. Returns video topics, post frequency, avg views, top hooks, hashtags used, and engagement patterns. Ideal for social media strategists, content teams, and competitive intelligence analysts.
- 安装量
- 2
- GitHub Stars
- 0
- 最近更新
- 5月13日